Understanding speech in a large crowd (such as a noisy room or cocktail party) remains to be one of the most challenging problems for hearing impaired subjects due to reverberation and multiple dynamic interferences. In some prior approaches, monaural or binaural microphone arrays have been used to improve speech understanding in such an environment. Due to reverberation and multiple dynamic interferences, the benefits have been limited in real-world situations. Monaural or binaural noise reduction algorithms have also been used to improve speech understanding in such scenarios. However, there is a need for improved speech understanding over what is currently available.